QIIME2 to Phyloseq Code from https://github.com/joey711/phyloseq/issues/821#issuecomment-371701469

1_qiime_part.sh

# ---
# title: Export QIIME2 OTU table to compatible file for phyloseq
# description: |
#   Three main steps to get to compatible file to import to phyloseq
#
#   Outline:
#   1. Export OTU table
#   2. Export taxonomy table
#   3. Export phylogenetic tree
# ---

# 1 Export OTU table
# - table-no-mitochondria-no-chloroplast.qza replace with your file
# - phyloseq => replace with where you'd like to output directory
qiime tools export \
  table-no-mitochondria-no-chloroplast.qza \
  --output-dir phyloseq

# OTU tables exports as feature-table.biom so convert to .tsv
# - Change -i and -o paths accordingly
biom convert \
  -i phyloseq/feature-table.biom \
  -o phyloseq/otu_table.txt \
  --to-tsv
  
# Manually change #OTUID to OTUID

# 2 Export taxonomy table
qiime tools export \
  taxonomy.qza \
  --output-dir phyloseq
  
# Manually change "feature ID" to "OTUID"

# 3 Export phylogenetic tree
qiime tools export \
  unrooted-tree.qza \
  --output-dir phyloseq
  
# 4 Merge files
# Filtered sequences will make taxonomy and OTU tables have different lengths

2_format_in_R.R

# ---
# title: Manipulate QIIME2 output in R
# description: |
#   Take in output from 1_qiime_part.sh and manipulate files in R
# ---

# Setup environment
library(here)

# Read in OTU table
in_path <- here("phyloseq", "otu_table.txt")
otu <- read.table(file = in_path, header = TRUE)
head(otu)

#	Read in taxonomy table
tax	<- read.table(Sile = "taxonomy.tsv", sep = '\t', header = TRUE)
head(tax)

# Merge files
merged_file <- merge(otu, tax, by.x = c("OTUID"), by.y = c("OTUID"))
head(merged_file)

# Note: the number of rows should equal your shortest file length, drops taxonomy
# for OTUs that don't exist in your OTU table

# Output merged .txt file
out_path <- here("phyloseq", "combined_otu_tax.tsv")
write.table(merged_file, file = out_path, sep = "\t", col.names = TRUE, row.names = FALSE)

3_import_to_phyloseq.R

# Setup environment
library(phyloseq)
library(ggplot2)
library(ape)

# Read in OTU table
otu_table_in <- read.csv("otu_matrix.csv", sep = ",", row.names = 1)
otu_table_in <- as.matrix(otu_table_in)

# Read in taxonomy
# Separated by kingdom, phylum, class, order, family, genus, species
taxonomy <- read.csv("taxonomy.csv", sep = ",", row.names = 1)
taxonomy <- as.matrix(taxonomy)

# Read in metadata
metadata <- read.table("metadata.txt", row.names = 1)

# Read in tree
phy_tree <- read_tree("tree.nwk")

# Import all as phyloseq objects
OTU <- otu_table(otu_table_in, taxa_are_rows = TRUE)
TAX <- tax_table(taxonomy)
META <- sample_data(metadata)

# Sanity checks for consistent OTU names
taxa_names(TAX)
taxa_names(OTU)
taxa_names(phy_tree)

# Same sample names
sample_names(OTU)
sample_names(META)

# Finally merge!
ps <- phyloseq(OTU, TAX, META, phy_tree)
ps

Hello!

I am trying to import the files from qiime2 ("moving pictures tutorial") to the phyloseq, but, during the step ## Import all as phyloseq objects, I got the error messages below:

1. For: OTU <- otu_table(otu, taxa_are_rows = TRUE) --> I got this message: Error in validObject(.Object) : invalid class “otu_table” object: OTU abundance data must have non-zero dimensions.

2. For TAX <- tax_table(taxonomy) --> I got this message: Error in dimnames(x) <- dn : length of 'dimnames' [2] not equal to array extent

All the previous steps worked. But I am stucked now.
Could you help me please?

Thank you for your attention

Regards